Dataset catalog

Structured datasets for real AutoTrain experiments

Review task fit, fields, source terms, quality checks, and a related workflow before importing a dataset into AutoTrain.

Catalog

Choose a dataset by learning objective

Classification

Iris

A classic 150-row, four-feature dataset for three-class Iris species classification.

Rows
150
Features
4 numeric
Task
Classification
Review dataset
Classification

Breast Cancer Wisconsin (Diagnostic)

569 diagnostic samples with 30 real-valued features for malignant-versus-benign classification.

Rows
569
Features
30 numeric
Task
Classification
Review dataset
Regression / classification

Student Performance

Student achievement records from two Portuguese schools with demographic, social, school, and grade attributes.

Rows
649
Features
30+
Task
Regression / classification
Review dataset
Multiclass classification

Obesity Levels

Eating-habit and physical-condition records labeled across seven obesity levels.

Rows
2,111
Features
16
Task
Multiclass classification
Review dataset
Regression / forecasting

LBMA Gold Price

Historical London Bullion Market Association gold price fixing data suitable for time-indexed regression experiments.

Frequency
Daily
Domain
Financial markets
Task
Regression / forecasting
Review dataset
Binary classification

Telecom Customer Churn Sample

A fictional telecommunications customer schema with service, account, billing, and churn fields.

Rows
7,043
Features
19 modeled
Task
Binary classification
Review dataset
Regression

California Housing

California census block features and median house values for regression experiments.

Rows
20,640
Features
8 numeric
Task
Regression
Review dataset
Clustering

Wholesale Customers

Annual spending across product categories for customers of a wholesale distributor.

Rows
440
Features
6 spend fields
Task
Clustering
Review dataset